i left my boat project for a while because it take a lot of time so i started to model a mixer
just on the same way i made more models. i am a newbie in Rhino so i learn the most by trial an error
but maybe its nice to see how i work and important to get feedback from you.
so will post much as possible screen shot with and without comment and hope the best.
first i look on the net to get some picture (now problem)
also look at turbosquit because you can find there good example models.
i start with a wire frame for the mean structure
i keep you informed at the process and my mistakes-
and i will ask a lot i hope not to much-
after a few day off trying now i need some help if someone like to help me - the first picture is
all the try i did bud no success at the end a found my way out with cage edit -
first picture 1
and at the end i made this but as you see the surfaces are not fitting very welt i think i try evrithing now but
still no good result 
i hope somebody can help me-
best.3dm (1.7 MB)
Hi Peter - Iāll be able to look more closely later, but as a start, I would never make the front part as you have from a single network surface - Iād say there are at least four basic surfaces here - two sides, the flat, (under the bowl), the curved-to-vertical front surface, plus some transition surfaces. As I say, Iāll dig in a bit more later on and see if I canāt make some more concrete suggestions.
@peter1, would approach this something like in the attached file -
best_PG_Example.3dm (260.9 KB)
Probably it is simpler than that - if there is a circular flat area, which seems likely, and not just the half circle, it may be that the curved surface is just a revolve. Something like this -
best_PG_Example2.3dm (429.4 KB)
-Pascal
Hi Pascal i will study on this way you did it my first impression is that the surfaces are more cleaner then mine
so i have mine doubts at the moment because shape of the body is more rounder and yours is a little tight or angled but mayby you fix this in a later stage ?? i have to keep in mind the picture of the mixer http://www.turbosquid.com/3d-models/stand-mixer-kitchen-aid-3d-model/538651 but i will learn your way to and look what happens - for so far thanks a lot !!!
Hi Peter- I was not modeling it for you, I was just showing how I would lay out the surfaces if I were making what I saw in the NetworkSrf version you posted. Modeling an existing object like this (in my opinion) it pays to spend some time analysing the shape and breaking it up insofar as possible into simple underlying shapes of relatively consistent curvature, and transition surfaces between them. That does not mean they all need to be as rigid as in my simple example, but I would avoid trying to make any one surface do ātoo much workā - that is, being a primary shape and a transition. NetowrkSrf encourages this in a way but that it not really what it is for - assuming you want a clean, well made model, I would use this command sparingly and not as a substitute for multiple surfaces.
BTW, from what I saw, I would also pay more attention to making simple and clean input curves - use the CurvatureGraph to helpā¦
Hope that $.02 worth helps someā¦ this looks like a fun model to make.
-Pascal
Pascal,
Could you elaborate a bit more on why you prefer multiple surfaces? I have noticed before that you often use more surfaces than those posting model problems, and you use matching a lot to get clean transitions.
I ask because I have always sought to use as few surfaces as possible. I do this for a couple of reasons:
- I use Flow to Surface a lot, and trying to evenly distribute objects over multiple surfaces is a headache.
- I like to preserve the option of untrimming holes, and they donāt untrim over multiple surfaces. Say there is raised lettering on the object and later we need to change the size or spacing or edit the wording.
- Fillets and champers have a habit of failing across multiple surfaces.
- Booleans donāt like it when surface edges happen to line up closely or wander back and forth across one another, they are cleaner with simpler intersections of the objects. Plus, if they donāt cross surface edges the original surfaces can be restored by extracting and untrimming if needed.
- Clean transitions are a pain (weakness in my skill set), and I often canāt get multiple surfaces to smoothly appear as one in the final polysurface.
- When the boss says āpull that face out and make the form fuller and softer around the lipā, adding a few profile curves and point editing some others is a lot quicker than having to re-create, match and join a half dozen complex transition surfaces.
- Maintaining history is a lot easier on a few editable surfaces than a bunch of small ones joined with non-history-enabled Matches and Fillets.
I guess if I knew exactly what the final object needed to be it wouldnāt matter, but nothing is ever right the first pass, and we adapt and edit parts constantly, so I put a very high priority on ways of working that permit later adjustments, even if the initial modeling takes longer.
I know this is partly personal preference (some people will spend hours reworking existing geometry while others would rather start fresh), but Iām sure you have good reasons for your style and Iād like to hear them.
Mark=
1 Like
hi pascal i was just working on it again and sorry it was not mine intention to say you are doing things wrong - you are
from a great help and i know now i have to keep it simple in the first place and your example help me on my way to find things out so in the evening i will post the results and try to carry on whit the model- there is a lot to learn-
after a few hoers the results are not good i am afraid so tomorrow i will start over again-
number1 was my first try but it gives ugly edge wen i mirror it alos te surface are not fitting good so i have to start over
number 2 the results are better but also problems wit fitting the surface together but its better i reed the CurvatureGraph
first i have ti find out what it means (i have to learn al this words because i am not a technical guy-
stupid thing is i made a lot off models and tey are pretty good but that was wen i not use my brains look for example
this coffee machine ?? dont ask me how i did it -
but here is the latest update 
and the file maybe i can can come further with it
greetings peterbest2.3dm (1.1 MB)
Hi Mark- Iām sure there are any number of good reasons to do things differentlyā¦ my impression is that Peter1 is asking for sort of āidealizedā modeling advice, and for what it is worth, I threw in my bitā¦ In general, I think you can have more control over the design if the surfaces are carefully arranged - you can certainly get away with not doing that in many cases but I think youāll find that surface models that are visually the most coherent tend to have carefully organized surface layouts. In particular, if the two directions of principal curvature are also close to the UV directions of the surface, that seems to be a good thing.
-Pascal
1 Like
Hi to all i am on my way yep pascal you are right i want to learn
1 i discover the best way it trying to reduce everything to box sphere cylinder etc planesā¦
i do this now on the top of the model unfortunately i like at the moment the bottom part the best from myself ā¦
however this is not good because the 2 surfaces wil not match front and the back
see picture 2 but here are the result til now and i am happy i know now more what i am doing in the top part of the mixer.
picture 1 
mixer5.3dm (1.4 MB)
i hope i can ask more here tomorrow i am conne do some simple things on the mixer i need some success-
greetings Peter
@peter1 Peter, do you have images of the back of this thing? Iāll try to helpā¦
-Pascal
i own one of those thingsā¦
could take some pics if you guys want.
That would be most excellent, even though it will probably make a mess of my guesswork so farā¦
-Pascal
no pascal you are doing well and help me a lot ā¦ but the bottom part red and bleu (yellow) thats mine problemā¦ i spent i think 2 days to get them fit together with you matchsrf method but but the red bleu surface will not going over in a smooth way??
thanks Jeff but the picture are no problem its just the modelling but thanks for the help maybe i will ask you wen i need some details if thats oke with you -
greetings peter
Hi Peter - if you Untrim the blue object and MatchSrf for Curvature to the red, you may see some improvement. However, making it as you are with a huge NeworkSrf will likely never be very clean. In my opinion, from a clean modeling perspective, you are spinning your wheels with this approachā¦
-Pascal
Hi Pascal- after this post i was a little disappointed but today i give it a try your method for the bottom part
and yes you are so right about it much much better 
now i have more control over it ā¦but if you can see the old wan has a little curve in the bottom part i dit this whit cage edit can i do this again wen i put all the surfaces together ??
greetings peter
by the way a little test render
Thatās the ticket. Looking much cleaner. I still think you can simplify things - probably no need, as far as defining the shape goes, for quite such dense surfaces, but that is the idea I would say. You can cage edit the curve into your object, but I would encourage building in that way from the start if possible.
-Pascal
this is how the back looks
Merci!
-Pascal
@jeff_hammond, could you please take a shot downward so I can see the footprint shape, and what it looks like where the bowl sits? It seems there are various similar ones out there, some squarish, some roundier. Arenāt you glad you offered??
-Pascal
some more views:
1 Like